Charles Don Flores is waiting for his execution. He has been in prison for more than 20 years after an eyewitness identified him as a perpetrator in front of a court in Texas. In the first and second comparison, however, she had not recognized him. His case is just an example for many similar malfunctions, a research team from the USA warns. Only the first comparison counts - after that, the memory is contaminated. In the case of Charles Don Flores, the statements of the witness therefore spoke for his innocence.
The group around the psychologist John Wixted from the University of California, including the well -known memory researcher Elizabeth Loftus, now wants to clarify this possible contamination. The focus is on the comparison, a process that lay people know primarily from US crime series: the investigators mix the suspect among several uninvolved people with whom he shares essential characteristics such as ethnic origin. Sometimes the people are in a row behind a disposable window, sometimes only photos of them are presented. The victim or an eyewitness should say whether he recognizes the perpetrator in one of the people.
There are already guidelines for comparison to avoid possible falsifications. For example, the witness should not have seen any media reports that could provide evidence of the appearance of the suspect. He has to be informed that the perpetrator may not be among the people. And the investigators who carry out must not even know which person is the suspect in order not to unconsciously influence the witness.
But even if the comparison goes according to all the rules of art, the result can be wrong, write Wixted and his team. »A comparison leaves a memory lane of all faces, including that of the suspect. As a result, the memory of his face will react more strongly in the next comparison, even if the person is innocent. ”The more often the witness sees the suspect, the greater the risk of false identification and the more he believes in remembering properly which is often considered an indication of a reliable statement in court. There is no way to "decontaminate" the memory again.
The Steve Titus case
In 1980, the Seattle police were looking for a rapist. Steve Titus, a restaurant manager, was stopped by the police on his way home at night because his car resembled the description of the perpetrator's vehicle. The victim only said during a first confrontation: "He looks the most similar to him." However, in court, the woman seemed to be sure of herself; her initial uncertainty was ignored and Titus was convicted. A few months later, the victim recognized the real culprit in a photo that appeared in the press in connection with another case. Titus was released, but died shortly after at the age of 35 from a heart attack.
The problem is known to science for a long time. As early as 1977, a study showed: When test subjects were presented to photos of supposed criminals and were to search them from a series of photos of innocent persons an hour and a half later saw first time. Recent studies come to similar results.
Co-author Elizabeth Loftus has become known for experiments in which she implanted such false memories in memory. Among other things, this happened with soldiers as part of a survival training: for 30 minutes they were subjected to aggressive interrogation. After that, they were supposed to report on the interrogation, and some of them were presented with a photo – supposedly from the person they were interrogated by. Later, they were to pick out the person they were interrogated by from a series of pictures. If the right person was not among them, more than half chose the person they had seen a photo of afterwards.
The group underpins the practical relevance of research with figures from the Innocence Project, a non -profit organization that is committed to the revision of judiciary. Accordingly, 70 percent false identities were jointly responsible for the malfunctions they have been informed, which were subsequently informed by DNA tests. In around 57 percent of these cases, the witness stated in court to be sure, although he had not been in the first comparison. Most of the time he even had nobody or someone else recognized as a perpetrator.
As early as 2017 and 2020, the group had therefore called for new guidelines for the questioning of eyewitnesses: "We have to prevent repeated comparison of the witness and suspects." This also applies to a comparison by picture. "Even a bad photo can contaminate the memory as long as the suspect can be recognized." On the other hand, a second comparison is not a problem if none of the people from the first round appear again.
Many errors in judgment would be preventable
The police report also had to say how certain the witness was in the first comparison. The entire procedure should also be recorded on video so that everyone can check that afterwards. Because if the witness is not sure, it is a warning signal - regardless of whether he expresses security in his own words or in percent. A slow reaction could also be a warning signal. Studies showed: A quick identification within 5 to 10 seconds was usually correct. If the decision lasted more than 30 seconds, it proved to be less reliable.
So what stands in the way of implementing the directives in practice? Wixted and his team write that the problem lies in the widespread misconception that later comparisons bring as much or even more information. Particularly problematic: identification in court. Because this situation is highly suggestive, since the suspect is already sitting in the dock.
The first test provides the only relevant evidence and the best chance of genuine memories, warn the authors around Wixted. "If this simple reform had been made years ago, many misjudgments would have been avoided."
